Proposal Summary Subpopulations of hematopoietic cells with different functions are identified based on the presence or absence of cell surface markers. Certain subpopulations are hypothesized to mediate forms of graft-vs-host disease (GVHD) in allogeneic hematopoietic stem cell transplantations (aHSCT). The most advanced efforts to minimize these side effects, and to have reached clinics, rely on an extended three step immunomagnetic separation protocol with sequential positive selection (for protection of stem cells from elimination based on CD34 expression), negative selection (elimination of nave TN-cells based on CD45RA expression), and subsequent return of positively selected stem cells to the graft that is then introduced into patients. The engineered graft, however, is also depleted of B-, NK-, and Treg-cells, which are essential for control of infectious complications, modulation of GVHD, and relapse control, thus impacting long-term survival after transplantation. Rapid and efficient single-step positive or negative selections that would minimize handling of cells out of their natural niches would also improve cell health likely leading to better clinical outcomes. Highl precision single-step procedures would also provide us with a powerful tool for rigorous and reliable studies of the impact of high quality preparations of specific subpopulations of cells in animal models of disease. We have recently developed molecular computing cascades based on oligonucleotides conjugated to targeting moieties (such as antibodies, their fragments, and aptamers). These are mixtures of targeting moieties that carry molecular computing elements and produce one outcome, a specific label on a single cell subpopulation, based on taking into account all targeting moieties. Thus, what would be, for example, a three-step separation protocol or a three-colored-fluorescence characterization in flow cytometry is condensed into a single step isolation procedure or a single color in flow cytometry. These computing mixtures of oligonucleotide conjugates with targeting moieties are thus uniquely suited for single-step `mix-and-separate' cell labelling protocols for magnetic separations. Our cascades can block isolation based on the presence of a cell surface marker or they can amplify signal coming from surface markers with low expression levels. In this project we will ask two principal questions: (1) What are the practical limitations, with metrics being yield, purity, and health of cells, of molecular computing cascades when applied to isolation of cell-subpopulations for clinical applications? And, (2) What are the minimal subpopulation of cells that we have to either preserve or eliminate to minimize induction of various forms of GVHD. In three Aims we will address translational and mechanistic questions regarding graft engineering and the induction of GVHD using increasingly complex protocols. We will characterize the impact of cell populations with different levels of markers for nave, central memory, and Treg cells in a mouse model of GVHD, while at the same time assessing translational potential of increasingly complex molecular computing cascades, in comparison to standard methods.